The role of the serotonin system in the etiology and pathogenesis of autism spectrum disorders (ASD) is not clearly defined. High levels of platelet serotonin (5-HT) have been consistently found in a proportion of patients, and it is known that specific 5-HT transporter gene (SLC6A4) variants modulate transporter reuptake function, therefore possibly influencing the occurrence of hyperserotonemia in a subset of autistic patients. We have examined the association of platelet serotonin levels with two SLC6A4 polymorphisms, 5-HTT gene-linked polymorphic region (HTTLPR) in the promoter and intron 2 variable number of tandem repeats (VNTR), in a sample of 105 ASD patients, their parents, and 52 control children. Quantitative transmission disequilibrium test (QTDT) results showed a significant effect on 5-HT levels of each SLC6A4 marker (P=0.017 for HTTLPR; P=0.047 for intron 2 VNTR) and of haplotypes of the two markers (P=0.017), with a major contribution of the L.Stin2.10 haplotype (P=0.0013). A 5-HT mean value in the range of hyperserotonemia was associated with the homozygous L.Stin2.10 haplotype (H (1,N=97)=7.76, P=0.0054), which occurred in 33% of hyperserotonemic patients against 6% of patients with normal 5-HT levels (Fisher's exact test: P=0.013, OR=8). Allele interaction at the HTTLPR locus was found, with a significant dominance variance effect on 5-HT levels. We found no transmission disequilibrium of any of the SLC6A4 variants in ASD. Our results show that the SLC6A4 gene is a significant factor in the determination of 5-HT levels, and that specific SLC6A4 variants are associated with an increased risk for hyperserotonemia in our sample of autistic patients. The biological mechanism, however, is unlikely to involve the SLC6A4 gene solely. The associated SLC6A4 alleles likely interact with other genes or environmental factors to produce the abnormally high 5-HT levels observed in this subset of autistic patients, who possibly represent a separate etiological group.